With half a month still left to go in the season, several first round playoff series nonetheless seem inevitable. Perhaps the most intriguing matchup of the batch is the Chicago Blackhawks and Colorado Avalanche, two teams in their first year of being division rivals.
The season series between the two went decidedly in Colorado’s favor, with the Avalanche winning four of five. As I have discussed in the past, however, individual game results can be misleading. This is especially true of small samples; one must keep in mind that goals are relatively random events that don’t necessarily dictate which team played better on a given night. Thus, as we analyze the imminent Chicago and Colorado playoff series, we will look at some trends established in those five games rather than taking the outcomes themselves as gospel.
Of course, this hardly avoids the sample problem. While a more nuanced look than the number of goals for and against is certainly the right way to go, there remains considerable room for variability. Such is the nature of trying to predict playoff success and failure in a league notorious for its parity. Nevertheless, because writing things off as too difficult to predict would be more boring than Bran Stark, we’ll take a shot at it anyway.
Shot totals are one of the more imperfect proxies for in-game control, but they’re fairly simple to look at and can be quite telling if they display a pattern.
And a pattern we have. Chicago and Colorado shots in each game, respectively: 37 to 23, 37 to 18, 48 to 26, 38 to 20, and 39 to 29, for an average of 39.8 to 23.2.
Double digit shot disparities occurring once or twice between two teams is one thing, but for it to transpire five times out of five in the same team’s favor suggests something more than randomness is at work. From Colorado’s perspective, the negative 16.6 differential would be twice as bad as the league’s worst rate at 5-on-5 (Toronto).
These numbers don’t seem to mesh at all with the results of the games themselves; as discussed earlier, the Avalanche won all but the second. This is just one example of score outcomes clearly being flawed as predictive mechanisms… not that we needed it.
Colorado’s wins against Chicago despite a terrible shot differential can be explained with one word.
In the five Blackhawks/Avalanche games, Semyon Varlamov has been about as brilliant as Corey Crawford and Antti Raanta have been horrid. Varlamov’s save percentage is a sparkling 96%; the Chicago goalies have combined for a dreadful 86%.
A common — but laughably faulty — interpretation of these numbers purports something along the lines of “Varlamov just has Chicago’s number, and the Avalanche have figured Crawford and Raanta out.”
Such fallacious thinking is rife in hockey circles; I’ve christened it “hockey speak,” a blanket term applying to any useless, vague cliches or epithets that talking heads or fans crow about and treat as ironclad evidence to one purpose or another.
The statement “X has Y’s number” in reality means the same thing as its much less nebulous sister statement, “X has played well against Y.” There is, of course, no guarantee or indication that X will continue to play well against Y.
Let us make the discussion more concrete, bringing Varlamov into the mix. Does his turning away 96% of Chicago shots in five regular season games constitute a reason to believe that he would maintain a similar or comparable percentage in a playoff series against those same Blackhawks?
No. Introductory statistics teaches us about averages and regression to the mean, and one does not simply ignore introductory statistics.
The same goes for the other example of “hockey speak” that I mentioned, that the Avalanche somehow “have Crawford’s number” (for playoff purposes, Raanta’s numbers are irrelevant). In a similar but opposite vein to Varlamov’s case, we would be best off hedging our bets on Crawford managing a save percentage higher than the pathetic 83% he compiled against Colorado in the five games this year.
In sum, the contrast of elite and horrific goaltending for the two teams probably made the difference between a 4-1 season series in favor of Colorado and a 4-1 season series in favor of Chicago. The performance that both teams received in net was almost certainly unsustainable, be it in the positive or negative direction.
Shot totals can sometimes be deceiving (although they are not in the this case). To further our analysis, let’s take a quick look at the Corsi data from the five games.
A quick explanation for those who don’t know: The stat I will be using is Corsi for percentage (CF%), which is a measure of what percentage of total shot attempts a team takes in a game. By including blocks and shots that miss the net, Corsi effectively eliminates the uncertainty inherent to looking at shot totals alone.
CF% data for the Blackhawks and Avalanche (respectively) in their five head-to-head matchups: 62 to 38, 66.3 to 33.7, 70.4 to 29.6, 64.4 to 35.6, and 55.5 to 44.5, for an average of 63.72 to 36.28 (stats via extraskater.com).
This is eye-popping stuff. For some context, Chicago’s CF% average against Colorado is six percentage points higher than the team leading the league in the stat, and the Avalanche’s paltry 36.28% would rank a whopping five percentage points lower than the league doormat Buffalo Sabres.
A good Corsi number has exhibited a strong positive correlation with playoff success, and the inverse is also true. This makes intuitive sense; Corsi is a proxy for possession, after all, and teams that dominate the puck tend to dominate games.
As for Chicago and Colorado, we are presented with another piece of evidence which suggests that the former is more likely to win a series between the two.
With the more interpretive stuff out of the way, we can conclude our consideration of the various relevant factors with something more concrete.
To start with, one prominent player presumably will not take part in this probable series.
Duchene out 3-4 weeks. Likely to miss first round
— Adrian Dater (@adater) March 31, 2014
Duchene is Colorado’s best player and most dynamic offensive weapon. Even with the deep, strong forward group the Avalanche have, he will be missed — especially in light of the scoring punch Chicago possesses.
The Blackhawks have some problems of their own on the injury front, although none quite as definite or concerning as the Avalanche’s. Patrick Kane’s grade 1 MCL sprain isn’t a grievous injury by any means, and he could even play through it if necessary (for comparison, Bryan Bickell sustained the more serious grade 2 MCL sprain during the 2013 playoffs and didn’t miss a game). Kane will reportedly be healthy and ready in time for the beginning of the playoffs. Still, his game is predicated on puck control, skill, and on-ice vision — three traits that are almost always a bit dulled by a long absence from game action. Rust will assuredly be in play.
Chicago’s other franchise face, Jonathan Toews, suffered an upper-body injury in a recent game against the Pittsburgh Penguins. His health status remains somewhat unclear as per the norm for the tight-lipped Blackhawks. Still, Joel Quenneville did state that he believes the injury is “not serious,” which we can reasonably take to mean that Toews will probably be in the lineup come playoff time.
Blackhawks Playoff Series vs. Avalanche: Prediction
Chicago has outplayed Colorado by a sizable margin in their games this season, lending credence to the position that game outcomes on their own don’t make for a strong predictive basis. Further, it is exceedingly unlikely that the two teams’ goaltenders will follow the pattern established during the regular season head-to-head matchups. Both teams have some injury questions, but Colorado has it worse; the absence of Matt Duchene is enormous.
Taking all factors into account, it would seem most wise to confidently pick the Blackhawks over the Avalanche in this all-but-inevitable first round playoff series. I don’t think home ice advantage will be especially important either way; it hasn’t been for several postseasons in a row.
There is, of course, always room for variability. Perhaps Varlamov does manage to maintain a sky-high save percentage throughout the series while Crawford stays down in the low 80s. I have not and will not claim that this is impossible; rather, it is simply improbable. Anything can happen in the playoffs, of course, but we are in the business of discussing what is most likely to happen.
Considering it all, then, my prediction is that Chicago takes the series in five games.